Digital Traceability for Reused Construction Materials: A Conceptual Protocol for Material-Flow Visibility Across Construction and Demolition Waste Stages

1. Introduction and research question

Reused construction materials are difficult to trace because the same material may shift from installed asset to waste stream, then to stockpiled reclaimed inventory, and later to a second use context. The key technical problem is not only quantifying how much material is present, but preserving a reliable identity link between the object observed in the field and the object recorded in a digital system. In this sense, traceability is a material-flow visibility problem: the system must show what the item is, where it is, in what condition it is, and which processing stage it has reached. Existing digital construction research offers several partial tools for this task. Automated waste recognition and volumetric estimation can observe objects on site, while material-processing analyses show that waste generation and cost effects differ across design, procurement, handling, storage, transport, and on-site management stages. Digital twin and BIM studies contribute the sensing, spatial registration, and information-management logic needed to connect physical observations with a structured digital record. However, these strands have rarely been assembled into a direct traceability protocol for reused materials. [1] [2] [3] [4] [5]

The research question is therefore: how can digital traceability for reused construction materials be defined as a formal, stage-aware information process that links waste quantification to material-flow control and supports reuse decisions without claiming empirical validation? This manuscript answers the question through a targeted narrative synthesis rather than a systematic review. The synthesis is intentionally selective and limited to methods that directly inform the traceability problem. It does not attempt to cover all sustainability applications, broader smart-city digital twins, or unrelated infrastructure monitoring. The resulting contribution is a conceptual protocol with explicit data fields, decision rules, and a prospective evaluation design. [1] [3] [5]

2. Evidence scope and synthesis logic

The evidence scope is narrow by design. The first stream concerns automated construction and demolition waste recognition, including image-based classification, 3D imaging, laser scanning, volumetric estimation, and the challenge of field variability. The second stream concerns the drivers of waste and cost across material-processing stages, which is important because traceability must attach to the stage where the material changes state or value. The third stream concerns digital twin and BIM methods that can register site objects in a structured model and support real-time or near-real-time decisions. The fourth stream concerns phased automation roadmaps, which are useful because traceability systems are rarely implemented as a single intervention; they require an ordered sequence in which information infrastructure precedes advanced automation. [1] [2] [3] [5]

From this evidence, three transferable propositions emerge. First, material identity can be observed through multimodal sensing, but the observation is incomplete unless linked to an information schema that persists across stages. Second, the most consequential failures in waste management arise when stage transitions are not logged, because then the system loses the relation between quantity, condition, and responsibility. Third, a digital backbone based on BIM or a digital twin is a prerequisite for traceability because it provides the spatial and semantic container in which recognition outputs can be stored and compared. These propositions are not empirical findings of this manuscript; they are the organizing assumptions of the proposed protocol. [1] [2] [3] [4] [5]

3. Conceptual model of traceability for reused materials

The proposed model treats each reusable material item or batch as a traceability object. A traceability object is defined by six mandatory fields: identifier, geometry or count, location, condition, stage, and provenance. Identifier refers to a persistent code assigned to a single item, pallet, batch, or homogeneous grouping. Geometry or count refers to the measurable quantity of the object, such as length, area, volume, mass, or discrete unit count. Location refers to the current spatial position in site coordinates or facility coordinates. Condition refers to reuse-relevant state, including intact, damaged, contaminated, partially dismantled, sorted, stored, refurbished, or ready for reuse. Stage refers to the material-processing state, for example installed, deconstructed, sorted, stored, processed, dispatched, or reinstalled. Provenance refers to the source asset, removal event, and time stamp of each state change. [2] [3] [1]

The operational rule for traceability completeness is as follows: for any traceability object i at time t, completeness C(i,t)=1 only if all six mandatory fields are non-missing and the most recent stage transition is supported by a timestamped observation; otherwise C(i,t)=0. A site-level traceability score can then be defined as the proportion of recorded objects with C(i,t)=1 at a given audit moment. This rule is intentionally strict because the purpose of traceability is not merely to store inventory records, but to maintain continuity between physical observation and decision record. A less strict rule may be used in implementation, but the strict version is needed for evaluation and falsification. [3] [4]

The model also distinguishes three levels of material flow visibility: object visibility, stage visibility, and decision visibility. Object visibility means the system can detect a material entity. Stage visibility means the system can classify the entity into a processing stage. Decision visibility means the system can show why a reuse or disposal decision was made, based on the recorded condition, quantity, and location. Existing digital twin work demonstrates object identification and spatial registration, while BIM-linked decision support and cost-oriented waste analysis suggest how object records may inform downstream planning. The present protocol links these levels into a single traceability chain. [3] [4] [2]

4. Proposed digital architecture

The proposed architecture has four layers. The sensing layer acquires field data through point clouds, images, or other site observations. The recognition layer classifies objects and estimates their measurable quantities. The integration layer maps recognized objects to BIM elements or digital twin entities. The event layer stores stage transitions and reuse-related decisions in an auditable log. This architecture follows the logic of previous digital twin site management work, in which point cloud semantic segmentation was used to identify object types and locations, and it is consistent with BIM as a repository for geometry, schedule, and cost information. The distinction here is that the data model is purpose-built for reused materials rather than general site safety or progress monitoring. [3] [4]

In operational terms, the recognition layer may combine three procedures. First, detect candidate objects from point clouds or images. Second, estimate quantity using geometric reconstruction or volumetric approximation. Third, assign a reuse-relevant class label using condition cues such as shape regularity, breakage, contamination, or stockpiling state. The protocol does not prescribe one model architecture; the prior literature includes deep learning and semantic segmentation for site objects, as well as image-based waste classification and 3D scanning for waste quantification. What is required is not a specific model family but a consistent translation from sensed object to recorded traceability object. [1] [3]

The integration layer should use a bidirectional mapping between the field object and the BIM or digital twin entity. One direction resolves a sensed item to the best-matching object class and location in the digital model. The reverse direction allows planners to query the model for reuse eligibility, remaining quantities, and storage status. This bidirectional mapping is necessary because reused materials are often neither fully standardized nor permanently fixed to a single status. The phase-based automation literature suggests that such a digital backbone should be established before more advanced optimization or mechanized reuse workflows are added. [5] [4]

5. Algorithmic decision sequence

The protocol can be expressed as a decision sequence. Step 1: acquire a site scan or image set and time-stamp it. Step 2: segment candidate material objects. Step 3: estimate measurable quantity and condition. Step 4: match the observation to an existing traceability object or create a new object if no match is sufficiently confident. Step 5: update the object stage and location. Step 6: write an event record containing observation source, confidence, and operator override if present. Step 7: compare the updated record against reuse rules and flag objects for recovery, processing, or disposal review. This sequence converts traceability from a passive register into an active control process. [3] [1] [4]

A practical matching rule is needed because reuse objects may be fragmented, stacked, or partially hidden. The protocol therefore proposes a conservative identity rule: an observed entity is linked to an existing traceability object only if spatial overlap, object class, and quantity similarity all exceed predefined thresholds. These thresholds are hypothetical and must be calibrated locally. If no existing object meets the rule, the system creates a provisional object with lower confidence status. This prevents silent merging of distinct items, which is especially important for recovered materials that can have similar shapes but different provenance or contamination histories. The threshold values are not claimed as general standards; they are placeholders for site-specific calibration. [3] [2]

The comparator for this protocol is manual or spreadsheet-based traceability, in which materials are listed as masses or batches without continuous identity across stages. Compared with that approach, the proposed system should be evaluated on its ability to reduce missing stage transitions, duplicate records, and untraceable stock. It should also be compared with a BIM-only control condition in which materials are represented geometrically but not linked to stage-event logs. This comparator matters because the literature shows that BIM can store rich project information, but it does not automatically solve real-time dynamic data capture or trace continuity. [3] [4]

6. Validation design and outcome measures

Because this is a protocol, the evaluation design is prospective. The recommended validation consists of paired physical and digital audits across matched material batches. For each batch, auditors would record the physical count or volume, the digital record of stage transitions, the number of detected reuse-relevant events, and any discrepancies between actual and recorded state. The audit should include at least one deconstruction or refurbishment context and one active construction context, because the visibility problem differs between installed assets and stockpiled materials. The aim is not to estimate national waste quantities, but to test whether the traceability chain remains intact under realistic operational conditions. [1] [2] [3]

The primary outcome measure is traceability fidelity, defined as the proportion of audited objects for which the physical state and the digital state agree on identifier, quantity class, location class, and stage. Secondary outcomes include temporal lag between observed physical change and digital update, false link rate, false split rate, and decision-usefulness score from expert reviewers. A failure case occurs when the system identifies an object but cannot maintain its identity after transfer, storage, or partial dismantling. Another failure case occurs when condition can be detected but cannot be translated into a reuse decision because stage semantics are not standardized. These failures delimit the method’s generalization. [3] [4]

The protocol should also test sensitivity to mixed materials, damaged items, occlusion, and irregular geometry. Prior work on waste quantification notes that dynamic, cluttered sites reduce algorithmic performance, and prior digital twin work shows that point cloud and semantic segmentation can identify objects but remain dependent on data quality. Therefore, the validation should report not only aggregate agreement, but also error patterns by object type and site condition. This is essential because reused materials often have more irregularity than newly delivered materials, which may reduce recognition accuracy and increase identity ambiguity. [1] [3]

7. Comparison with prior methods and boundary conditions

The proposed protocol differs from automated waste quantification studies in scope. Those studies primarily address recognition, sorting, and quantity estimation of construction and demolition waste, whereas the present method extends the information chain to reuse eligibility, provenance, and stage continuity. It also differs from site-level digital twin studies that focus on safety management or object localization, because the target outcome here is not hazard detection but material-flow visibility. At the same time, the protocol relies on those studies for the recognition and registration primitives that make traceability feasible in a dynamic site environment. [1] [3]

The protocol also differs from stage-based waste-driver modeling. Structural equation modeling of material waste drivers shows that handling, storage, and transportation are especially consequential for cost overrun, which supports the decision to treat stage transitions as critical traceability events. However, that literature explains why waste occurs; it does not specify how digital records should preserve material identity across those stages. The present manuscript uses stage-based evidence as a design constraint, not as a direct model of reuse behavior. [2]

Boundary conditions are explicit. The protocol is most suitable where the material stream is discrete enough to be tagged or batch-labeled, where a BIM or digital twin environment already exists or can be established, and where site teams can perform repeated scans or image capture. It is less suitable for highly fragmented debris, chemically degraded material that no longer has stable reuse value, or projects without a persistent information backbone. The phased automation roadmap literature suggests that digital infrastructure should precede sophisticated automation; this applies here as a condition for implementation rather than a claim of universal applicability. [5] [4]

8. Conclusion

This manuscript establishes a conceptual basis for digital traceability of reused construction materials as a stage-aware, object-level information problem. It identifies the minimum data fields required for a traceability object, formulates a strict completeness rule, and proposes a decision sequence that links sensing, recognition, BIM or digital twin integration, and event logging. The literature supports the view that automated recognition, volumetric estimation, and digital twin registration are feasible building blocks, while stage-based waste analysis and phased automation planning clarify where the system is most likely to fail or succeed. [1] [2] [3] [5]

What remains untested is whether the proposed traceability chain can sustain identity continuity under real reuse conditions, especially when materials are damaged, mixed, hidden, or transferred across organizational boundaries. The protocol therefore invites validation against paired physical and digital audits, with explicit measurement of fidelity, lag, and error modes. Its contribution is bounded: it does not claim to solve reuse logistics in general, but it does provide a technically explicit framework for testing whether reused materials can be traced from waste generation to second-use entry with measurable visibility. [4] [3] [2]

References

1. Haq, Moinul; Mohsen, Osama; Elnabwy, Mohamed T.; Martinez, Pablo (2026). Integrating automation into construction waste quantification: A review of digital innovations for sustainable resource management. Waste Management & Research: The Journal for a Sustainable Circular Economy. 44(9), 1308-1326. doi:10.1177/0734242x261441094.
2. Kamal, Lobna; Yamany, Mohamed S.; Hosny, Hossam E. (2026). Assessing the impact of construction material waste drivers on project cost overrun across material processing stages: a structural equation modeling approach. Scientific Reports. 16(1), 28364. doi:10.1038/s41598-026-63008-x.
3. Qin, Wenli (2026). Application of construction site management based on digital twin and point cloud semantic segmentation technology. PLOS One. 21(3), e0340274. doi:10.1371/journal.pone.0340274.
4. Albtoush, Faten; Al-Btoosh, Ja’far A. Aldiabat; Rawashdeh, Taiseer Mustafa; Al-Btoush, Mohammed A. K. A. (2026). Assessing the impact of AI-enabled BIM digital twins on construction risks: a Bayesian Network of Jordanian expert beliefs. Frontiers in Artificial Intelligence. 9, 1846388. doi:10.3389/frai.2026.1846388.
5. Musyafa, Albani (2026). A strategic roadmap for construction automation in indonesian mass housing projects. Scientific Reports. 16(1), 15419. doi:10.1038/s41598-026-46773-7.